Research On Degradation Of Proton Exchange Membrane Fuel Cells For Vehicle Under Temperature Influence

Proton exchange membrane fuel cells(PEMFCs)are receiving increasing attention recently due to their characteristics of high energy conversion efficiency,good safety and zeroemission,which have broad application prospects in hybrid vehicles.The typical operating conditions of fuel cell hybrid vehicles(FCHVs)in the process of running such as start-stop,acceleration,and speed change will lead to drastic changes in the temperature,which greatly affect the output performance and the remaining useful life(RUL)of PEMFC.The optimum operating temperature range of PEMFC is 333K-353 K,and the thermal management strategy(TMS)is an effective way to improve the durability and operating efficiency of the PEMFC stack.An FCHV model from Autonomie software is selected in this research,where the rulebased management strategy is chosen to allocate the energy.In thermal management research,a novel machine learning(ML)algorithm is applied to the TMS of a PEMFC stack to control the inlet and outlet temperatures,and the proton exchange membrane(PEM)water content is also kept at a reasonable level.The control results of the proposed TMS are compared with three other machine learning regression algorithms and fuzzy control respectively.Results show that the proposed TMS keeps the inlet and outlet temperatures at about 338 K and 343 K respectively and has the best performance in reducing the maximum deviation of inlet and outlet temperatures and the fluctuation range of PEM water content,which improves the control accuracy of thermal management.In addition,the degradation of PEMFC is researched based on the proposed TMS by considering the typical working conditions of the FCHVs.To solve the problem of insufficient accuracy of traditional prediction methods of the PEMFC RUL,a novel nonlinear degradation model based on a first-order control system is adopted,where the vehicle operating conditions are divided into start-stop,idling cycle,high power load cycle and load changing cycle,and the PEMFC RUL is characterized by the variation of output voltage.The nonlinear degradation model describes the voltage variation of PEMFC in the later period of operation and the voltage degradation mechanism of PEMFC is explained better.The proposed ML-based TMS of the PEMFC stack in this research solves the problem of low control accuracy and robustness of traditional control methods and shows the best control performance compared with other thermal management strategies.The proposed nonlinear model-based method can accurately predict the RUL of the PEMFC stack under typical operating conditions.The methods and research ideas mentioned in this research have made exploration for the thermal management and life evaluation of the PEMFC stack.